Trochlear Morphology Development: Study of Normal Pediatric Knee MRIs

Introduction:

Trochlear dysplasia can significantly contribute to the development of patellofemoral instability. Trochlear dysplasia has been defined as a sulcus angle of greater than 145 degrees^1,2 and can be due to either decreased lateral trochlea slope of the lateral trochlea facet or decreased central trochlear depth³. The Dejour classification system has been used to describe the final, adult morphology of the dysplastic trochlear groove on radiographs⁴ yet there is no classification system for and very little understanding of the skeletally immature trochlea.

Despite an increased understanding of abnormal pathology, there is incomplete understanding of normal osteo-cartilaginous development of the trochlea. Nietosvaara et al studied normal patellofemoral anatomy using ultrasound and showed the cartilaginous trochlear sulcus is well developed and nearly in adult form at birth⁵. However, the underlying subchondral bone was not congruent with the overlying cartilage. They hypothesized that the subchondral bone grew and developed into a trochlear groove with age. A CT scan study of 31 pediatric knee specimens age 2–11 years showed a linear increase in medial and lateral trochlear height with age with a decrease in osseous sulcus angle, cartilaginous sulcus angle and patella sulcus angle until age 8 years⁶. The study findings are limited, however, based on small number of cadavers involved and the limited ability of CT scan to evaluate cartilaginous anatomy.

As knowledge increases of trochlear dysplasia and how abnormal anatomy contributes to patellofemoral pathology, understanding normal trochlear development is fundamental. Cartilaginous and osseous development of the trochlea has not been evaluated in a large sample size or by advanced imaging techniques, which best differentiates these two anatomical parts. This purpose of this study is to define the morphologic development of and determine the relationship between trochlear osseous and cartilaginous anatomy in pediatric patients using MRI evaluation.

Methods:

Prior to study initiation, IRB approval was obtained #18–001461-AM-00001. A retrospective chart review of knee Magnetic Resonance Imaging (MRI) from 2009 to 2017 was completed. Inclusion criteria were age 3– 16 years without significant history, exam, or radiologic evidence of patellofemoral pathology. Exclusion criteria included any abnormal findings on MRI image or radiology report, a history of patella instability or extensor mechanism injury, history of patella or femoral osseous abnormality including infection or tumor, congenital diagnosis that could affect bone and cartilage development, or any other diagnosis possibly affecting the knee development. Gender, age at MRI, and height at date of MRI were recorded to analyze for significant correlation with patellofemoral joint (PFJ) measurements. A chart review of orthopedic notes, imaging and radiology report confirmed inclusion. Two study authors (M.T. and B.K.), both orthopedic residents, completed measurements independently to obtain inter-observer reliability. In the study design process, the authors trained together and reviewed measurement techniques with the PI, attending orthopedic surgeon prior to analyzing the 246 patient MRIs. An equal distribution of sexes and ages were included. The power analysis was based on using a two-sided alpha level of 0.05 for having 90% power to detect a difference between sexes in the Pearson correlation of age and trochlear measurements. This was not specific for a particular measure. The required effect size was computed for a minimum sample size N=60, which was smaller than the final 246 knees analyzed. Under an assumption that age would show a fairly strong correlation (~0.7) in at least one of the sexes, the minimal effect size required in the second group to detect a difference from this correlation was 0.9.

Trochlea measurements were made on axial MRI imaging through the most developed portion of the trochlea. For consistency in choosing the axial cut for measurements, axial, coronal and sagittal images were cross referenced to find the image between the distal femoral physis and intercondylar notch with the most concave trochlea. The axial slice was selected by determining that in which the anterior to posterior dimensions of the condyles were the largest visible throughout the images of the distal femur. The selected axial slice also was required to include clearly defined posterior condyles, as their tangent is used as the reference line for the condyle height measurements. There was no specific slice thickness utilized for inclusion.

Measurements of the trochlea were determined by previously established methods ^3,7. Measurement of the trochlear groove included six height measurements perpendicular to the posterior condylar tangent line. These included from the edges of the articular cartilage (AC) and from the edge of the subchondral bone (SCB) to the posterior condylar reference line. Six individual height measurements were completed: maximum height (distance from the posterior condylar line) of the medial trochlea (MTH), lateral trochlea (LTH)) and the central trochlea (CTH) were recorded for the AC and for the SCB surfaces (Figure 1). Six individual angles were measured: the sulcus angle (SA) (Figure 1), lateral trochlear slope (LTS) and medial trochlear slope (MTS) of the AC and SCB (Figure 2).

Figure 1: Sulcus angles and trochlear height measurements on axial MRI.

Lines depict the measurements for the selected cut of each knee analyzed.

Sulcus angle measured along the edge of the articular cartilage and sulcus angle measured along the edge of the subchondral bone.

Trochlear heights: lateral, central, and medial heights measured from the posterior condylar reference line to the edge of trochlea for both sub-chondral bone and articular cartilage outlines.

Figure 2: Trochlear slope measurements on axial MRI.

Lines depict the measurements for the selected cut of each knee analyzed.

Trochlear slopes of the medial and lateral edge of both articular cartilage and subchondral bone (measured angle from the posterior condylar reference line).

All measurements are included in a comprehensive table, recorded by age quartiles, as a supplemental table (Table 2) for further review.

Statistical analysis:

Inter-rater reliability was performed using an interclass correlation coefficient (ICC) for absolute agreement to the average of the two reviewers scores, using a mixed effects model treating the reviewer rater differences as fixed effects. Interrater agreement was high across all measures with ICC values >0.7. Analysis was conducted using multivariate regression and linear regression analysis (quadratic model for age), with Huber-White adjusted standard errors for patient level clustering. Additionally, a quadratic model (age + age squared) was used to approximate the age at which the slope of the age-trochlea association changes based on the first derivative from the quadratic model.

Results:

246 knee MRIs from 230 patients were included in this study. 113 patients (51%) were female while 117 (49%) were male. 116 MRIs (47%) were of the Left knee and 130 (53%) were Right knee. Average patient age was 11.4±3.4 years. Interrater agreement was high (ICC values >0.7). To analyze mean values among age groups, the patients were broken down into quartiles by age, with at least 60 patients in each quartile. Mean values for measurements are presented by age quartiles (Table 1).

Table 1:

Trochlea Sulcus Angles (AC & SCB) across age quartiles

Sulcus Angle (SA)	1st Quartile (age 5.1 – 8.3) (N) Mean ± SD	2nd Quartile (age 8.3 – 11.5) (N) Mean ± SD	3rd Quartile (age 11.5 – 14.3) (N) Mean ± SD	4th Quartile (age 14.3 – 16.9) (N) Mean ± SD
Sulcus Angle (AC)	(61) 145 ± 6	(60) 142 ± 6	(60) 143 ± 6	(60) 145 ± 7
Sulcus Angle (SCB)*	(61) 151 ± 9	(60) 142 ± 6	(60) 137 ± 7	(60) 139 ± 8

^*change in SCB SA over ages was statistically significant (p<0.001)

LTH, MTH, and CTH showed linear increase with age (range 2 to 2.6 mm per year, p<0.001) (Figure 3). SA, LTS measured at the AC showed no change with age (p>0.05) (Figure 4). LTS and MTS measured at SCB showed significant increases with age (0.6 and 0.9 degrees per year, p<0.001), while SA for SCB showed a decrease with age (−1.4 degrees per year, p<0.001). There were no significant differences found in the age associations by laterality.

Figure 3:

Graph of Trochlear heights vs Age, showing consistent increase in heights with increasing age

Figure 4:

Graph of Trochlea sulcus angle of the articular cartilage and medial and lateral trochlear slopes of the articular cartilage, showing minimal change with increasing age

There were no sex differences in the age associations for SA, LTS, MTS (p>0.05) in either AC or SCB measurements. However for MTH, LTH, and CTH, for both AC and SCB measurements were found to have a significantly greater growth rate (p<0.001). Males had larger trochlear dimensions per additional inch of overall height when compared to females. Analysis showed that the LTS-AC had no statistically significant change over sexes or age quartiles. However, LTS-SCB, MTS-AC and MTS-SCB all showed significant increase over age quartiles (p<0.026).

The average articular cartilage slopes and sulcus angle plateaued at 11 years for both sexes. The average subchondral bone slopes and sulcus angle plateaued at age 13 for both sexes, indicating final morphology development. (Figure 5) Height measurements continued to increase linearly throughout the study age groups.

Figure 5:

Graph of Trochlea Sulcus Angle of the Articular Cartilage and Subchondral Bone vs Age, showing plateau at age 11 and 13 years respectively.

Discussion:

Analysis of 246 normal knee MRIs of a pediatric population ages 3 – 16 years demonstrated a consistent pattern of development of the trochlea. There was an increase in articular cartilage and subchondral bone medial, lateral, and central trochlear heights over time, consistent with overall growth of the knee. In addition, there was an increase in subchondral bone lateral trochlear slope and medial trochlear slope, with corresponding decrease in sulcus angle. However, no significant change in articular cartilage sulcus angle and lateral trochlear slope was found. This normative data indicates that the cartilagenous anatomy represents the mature trochlear shape at an earlier age than does the underlying subchondral bone contour. These results are similar to a previous MRI study of adolescent age dysplastic trochleas that reported no change in sulcus angle with age, and that trochlear shape, even in dysplasia, is largely predetermined by genetics.⁸ Together these studies confirm that articular cartilage morphology in patients ages 3–16 years demonstrates the mature trochlear shape early while the subchondral bone continues to change in shape below the articular surface during later development.

Studies in both non-weight bearing and 25% weight bearing conditions have shown throughout flexion that the lateral trochlear slope has a significant effect on patella alignment.⁹ With quadriceps contraction, the lateral anterior femoral condyle has more contact with the patella than does the medial facet.¹⁰ In accordance with the importance of the lateral trochlea in patella tracking, this study found the lateral trochlea to be most consistent anatomical determinant of trochlear morphology. Despite consistency of lateral trochlea development, the articular cartilage sulcus angle changed minimally over age quartiles due to minor changes seen in the medial trochlear slope articular cartilage. Therefore changes of the sulcus angle with age were minimized by the consistency of the lateral trochlea shape. The intimate developmental and mechanical relationship between the patella and lateral trochlea is paramount in understanding patellofemoral instability.

Previous studies have also reported minimal change in sulcus angle, consistent from early development to adulthood. Glard et. al used a biometric protocol originally described by Wanner¹¹ to compare the femoral patellar groove of fetus cadaver knees to previously reported adult knees showing a minimal difference in the sulcus angles and trochlea slope angles from fetus to adult. They concluded that the sulcus angle was largely predetermined by genetics.¹² Although these studies separately describe young and mature trochlea, neither study analyzed longitudinal development over time. This current study provides data of trochlea development from age 3–16 years.

In regards to height, the six trochlear height measurements, including both AC and SCB, increased linearly with age. Predictably the height differences between sex became more notable after age 10.

Similar to DDH interventions performed to keep the femoral head centered within the acetabulum, so that force of the femoral head on the soft articular cartilage contours a new acetabulum, it can be theorized that re-centering a subluxated or dislocated patella in the trochlear groove would guide development of a more concave trochlea and improve stability of the patellofemoral joint. Ideally intervening at a younger age will increase the effect and correction. Despite their intimate association, few studies have evaluated the relationship of the developing patella on the developing trochlea. Wang et al studied abnormal patellofemoral anatomy by simulating patella subluxation in rabbits and showed that the patellar articulation with the trochlea affects the development of this joint¹⁵. This study has not been expanded beyond an animal model, but it affirms a possible relationship between the patella and trochlea development. In 2016 Sugimoto et al reported on two cases in which they saw radiographic remodeling and improvement of trochlear groove angles after soft-tissue patella re-alignment surgery.¹⁶ They performed a Roux-Goldthwait procedure for an 11 year-old female and a medial reefing procedure for a 12 year-old male to address symptomatic, recurrent patella-femoral instability that had failed conservative therapies. Congruence angles pre and post corrective surgery of both cases demonstrated radiographic improvement in addition to clinical success of decreased pain and instability at 1 year follow up. They concluded that corrective surgery in the pediatric population has the potential to impart permanent anatomic improvement due to bony remodeling in an age where the trochlea is still developing. Currently, early centralization of the patella via various soft tissue techniques is the primary surgical option in skeletally immature patients. This study shows plateauing of the trochlear morphology by age 11 years indicating correction of patellofemoral tracking performed prior to this age may affect final trochlear morphology. With improved understanding of trochlea development, discussion of surgical intervention can be entertained, however timing and techniques are still debatable

Limitations of the study include variations in MRI cuts and the quality of the images, which were not standardized. MRIs were included from multiple imaging centers, and imaging protocols were not identical across all MRIs that were analyzed. At times only one axial series was provided, therefore measurements were made off of T1 & T2 sequences, depending on availability, and which had the highest quality imaging of our critical landmarks, ie condylar outlines. Variability in slice thickness and/or knee flexion during patient position may have also introduced systematic error. Furthermore, the anatomy of the trochlea curves posteriorly as it becomes the intercondylar notch distally. Therefore, when analyzing the groove the oblique detail is missed on distal standard axial cuts of an MRI. The MRI cuts analyzed were proximal to the intercondylar notch to avoid this aspect of the anatomy. However based on the MRI cuts taken, the ideal image was not always available, and the closest cut that still had clear posterior condyles was utilized. Differences in the sequence protocol, the imaging center, patient positioning, and the age due to the size of the knee imaged, made the cuts available inconsistent and adds variability in our measurements.

Another limitation was the lack of ethnicity recorded in the data set demographics. Therefore the opportunity to analyze for variations due to genetic skeletal differences among ethnic groups was missed. Furthermore, although trochlea dysplasia was excluded from our data set, there were likely other pathologies, without recorded diagnoses on chart review, which may have affected knee anatomy, but were not analyzed for effect on the PFJ in our study.

Conclusions:

This study found an increase in articular cartilage and subchondral bone medial trochlea height, lateral trochlea height, and central trochlea height over time as well as an increase in subchondral bone lateral trochlear slope and medial trochlear slope, with decrease in sulcus angle. However, the articular cartilage of the lateral trochlear slope and sulcus angle remained constant, with no significant change throughout growth. This normative data indicates that, the lateral trochlea slope and sulcus angle of the articular cartilage are predictors of final trochlea shape in normal development. Final trochlear morphological development is nearly complete around age 11, for articular cartilage and age 13 for subchondral bone. After age 11, the articular cartilage shows no significant morphological changes, other than overall linear growth occurring thereafter. After age 13, the subchondral bones matches the final shape and morphology of the overlying articular cartilage and likewise only grows linearly, maintaining anatomical shape, without any changes in angular morphology.

Supplementary Material

Supplemental Data File (.doc, .tif, pdf, etc.)

Supplemental Table of all trochlear measurements by age:

Table 2: Descriptions of Trochlea measures across age quartiles